FI3878175T3 - Quantization for video encoding or decoding based on the surface of a block - Google Patents

Claims (15)

Patent Claims 1. A method for video coding which calculates: - when a block of an image is divided (504) into sub-blocks: - I attach (505) a subdivision level counter to the sub-block; - determining the quantization parameter for the subblock based on the counter of the subdivision level; and - encode the subblock based on the specified quantization parameter, where the subdivision level counter represents the size of the subblock and is incremented (403, 405, 407) by a value representing the type of subblock division when the image block is divided into subblocks price 2. A method for decoding a video, which method comprises: - when the image block is divided (504) into sub-blocks: - a subdivision level counter is connected (505) to the sub-block; —- determine the quantization parameter for the sub-block based on the sub-section level counter; and - decode the subblock based on the specified quantization parameter, where the subblock level counter represents the size of the subblock and is incremented (403, 405, 407) by a value representing the type of subblock division when the image block is divided into subblocks price 3. Device (1000) for video coding, comprising an encoder (1030) configured: - when a block of an image is divided into sub-blocks: — - to connect a subdivision level counter to the sub-block; - to determine the quantization parameter for the sub-block based on the counter of the sub-division level; and - to encode the sub-block based on the specified quantization parameter, where the sub-block counter represents the size of the sub-block and is incremented by a value which - represents the type of sub-block division when dividing the image block into sub-blocks. 4. Apparatus for video decoding, comprising a decoder (1030) configured: - when the image block is divided into subblocks: - to connect the subblock level counter to the subblock; - to determine the quantization parameter for the sub-block based on the counter of the sub-division level; and - decoding the sub-block based on the specified quantization parameter, where the sub-block counter represents the size of the sub-block and is incremented by a value which - represents the type of sub-block division when dividing the image block into sub-blocks. 5. A method according to one of claims 1 or 2 or a device according to one of patent claims 3 or 4, which further comprises: - allowing a new quantization group according to the type of division, when the sub-division level counter of the sub-block is smaller than or equal to the maximum value of the subdivision. 6. The method or device according to claim 5, which further comprises: - when a new quantization group is in use, recording the position of the block (503). 7. A method or device according to one of claims 5 or 6, which further comprises: - disabling new quantization groups when the image block is divided into sub-blocks according to the three-part division and when the sub-division level counter of the smallest sub-block is greater than the maximum value of the sub-division. 8. A method according to one of claims 1, 2 or 5-7, or a device according to one of patent claims 3-7, in which the increment value is equal to: two for all subblocks of four division, one for all subblocks of binary division, — two for side blocks of three-part division and one for three-part division to the middle subblock. 9. A method according to one of claims 1, 2 or 5-7 or a device according to one of patent claims 3-7, in which the increment value is equal to: — four for all sub-blocks of the four division, two for all sub-blocks of the binary division, four for the side sub-blocks and two for the central sub-block of the ternary division, four for the smallest subblock and one for the largest subblock in an asymmetric four-part binary division, and three for the smallest subblock and one for the largest subblock in an asymmetric one-third binary division. 10. A method according to one of claims 1, 2 or 5-9 or a device according to one of patent claims 3-9, where the quantization applies to fog components or color components. 11. A recording medium in which video bitstream data is encoded and which comprises at least a syntax data element that represents a value representing a counter of the subblock level, which — represents the surface of a subblock, where the subblock is the result of dividing a block of the video image and the counter of the subblock level is increased by a value that represents the division type of the subblock, when an image block is partitioned into sub-blocks. 12. A recording medium according to claim 11, where the increment value is equal to: two for all subblocks of the four-part division, one for all sub-blocks of the binary division, two for the side blocks of the three-part division and one for the central sub-block of the three-part division. 13. A recording medium according to claim 11, in which the increment value is equal to: four for all subblocks of the four division, two for all subblocks of the binary division, — four for the side subblocks and two for the subblock of the three division, four for the smallest subblock of the asymmetric quarter binary division and one for the largest subblock, and three for the smallest subblock and one for the largest subblock in an asymmetric one-third binary division. 14. A computer program that contains program code instructions to be executed by the processor (1010) to implement the steps of at least one of the methods according to claims 1, 2 or 5-10. 15. A computer program product that is permanently stored on a computer-readable data medium and contains program code instructions which, when executed by the processor (1010), enable the device (1000) to implement the methods according to at least one of claims 1, 2 or 5-10.